Diabetic Retinopathy Severity Classification Using Multi-Label Support Vector Machine

This final project aims to classify the severity levels of diabetic retinopathy based on the occurrence of segmented lesions. This project utilizes the IDRID, DRIVE, and STARE datasets for the segmentation and classification processes, and also build GUI with streamlit.

Insul.in : Early Diabetes Detection App

With my team, we developed an application for early diabetes detection. This application uses kaggle dataset for learning and deep learning model to make predictions. In addition, there are many useful features for user.

MNIST Handwriting Classification Using ANN & CNN

This project aims to classify MNIST handwriting dataset containing images with 0-9. After build the model, I use streamlit as user interface for the prediction.

Medical Image Segmentation and Labelling

This project focuses on knee MRI image segmentation and labelling its segmented objects. For further development, the labelling result could be for disease detection/prediction.

Machine Learning Algorithm for Classifying Diabetic Retinopathy Severity

This project is an extended version of my final project, using many types of existing machine learning algorithm to classify diabetic retinopathy severity, such as : Support Vector Machine (SVM), Random Forest, K-Nearest Neighbors (KNN), Naive Bayes, Decision Trees, Logistic Regression, etc.

Restaurant Recommendation System Using Cosine Similarity

In this project, I developed a sophisticated restaurant recommendation system employing Natural Language Processing (NLP) techniques to enhance user experience and satisfaction. Leveraging the power of cosine similarity, I engineered an intelligent solution that seamlessly matches user preferences.

Bidirectional LSTM Model for Text Generation

In this project, I engineered a robust Bidirectional Long Short-Term Memory (LSTM) language model tailored for text generation, focusing on the rich linguistic nuances of Irish lyrics. Leveraging the power of Natural Language Processing (NLP), I harnessed bidirectional architecture to create a model

Exploring SpaceX Dataset through Data Science

The aim of this project is to examine the data from SpaceX Falcon 9 that has been gathered from diverse origins and utilize Machine Learning algorithms to forecast the outcome of the initial stage landing. This prediction allows other space organisations to determine whether to compete with SpaceX

Heart Rate Detection

In the QRS detection process for heart rate determination, pre-processing and segmentation steps are performed to reduce noise and extract PQRST waves from a single ECG cycle. Subsequently, the Pan Tompkins QRS Detection Algorithm is applied.

Synthetic ECG Generation

The Synthetic ECG Generation project creates artificial ECG signals by manipulating parameters such as frequency, std, and PQRST parameters, offering diverse patterns for valuable insights in research and training.

Netflix Binge-Watching Recommendation System

Netflix Recommendation: Using EDA & cosine similarity by genre, this system offers personalized movie suggestions, enhancing user satisfaction.